Chapter 24 of the Book of Zambasta : A New Version of the Prophecy of the End of Buddhism in the Candragarbha sūtra

: The Book of Zambasta is a Khotanese Buddhist work that collects various Buddhist texts; and its last chapter, Chapter 24, contains a so-called famie 法滅 prophecy, that is, the prophecy about the end of the dharma. This chapter has been identified by Jan Nattier as corresponding to the ‘Fa miejin pin’ 法滅盡品 [Chapter on the Complete Disappearance of the Dharma] of the Yuezang jing 月 藏經 [Skt. Candragarbha sūtra ; Bodhisattva Candragarbha Sūtra]. In this article, I compare Chapter 24 with three other versions of the prophecy from the so-called Candragarbha sūtra group. My aim is to highlight five major points of discrepancy, with a particular emphasis on the battle scene in Chapter 24 which, in my opinion, is unusually elaborate. This battle scene, I venture to speculate, may reflect certain influences from the Persian epic Shahnameh and, more probably, the Indian epic Mahābhārata . By comparing these epics with the Book , I raise the possibility that perhaps non-Buddhist literary traditions have played some roles in shaping the Buddhist famie prophecy. Furthermore, I argue that Chapter 24 is part of a non-extant version of the Candragarbha sūtra . My attention then turns to the Candragar-bha sūtra itself, as I discuss its geographical origin (i.e., Khotanese) and the date (early sixth century). The article ends with a discussion on the date of the Book of Zambasta . I especially focus on showing that Mauro Maggi’s dating of the Book , i.e., the second half of the fifth century, is too early.


Book of Zambasta
T he Book of Zambasta (hereafter, the Book) is a Khotanese Buddhist work that collects an array of Buddhist sources. 1 It is estimated that it originally comprised 298 folios, among which 208 have been discovered and studied. They are preserved in the St. Petersburg Branch of the Institute of Oriental Studies in the Russian Academy of Sciences, British Library, Yale University, Museum of Indian Art in Berlin, Asiatic Society of Bengal in Calcutta, and Ryukoku University in Kyoto etc. The majority of the extant folios, 192 in total, were collected by the Russian consul-general Nikolai Petrovsky  in Xinjiang. The Book has been translated into several modern languages, with the most authoritative transcription and translation being the 1968 English translation made by R. E. Emmerick (1937Emmerick ( -2001. 2 During the decades following Emmerick's translation, however, the interests in the Book had abated. In recent years, some scholars have revived the interest in the Book. For instance, the traditional opinion was only able to determine that the Book could not be composed earlier than the seventh century, but Mauro Maggi proposes that the Book could date to as early as the mid-fifth century (though I will challenge his dating later). 3 3 Maggi, 'The Manuscript T Ⅲ S 16', 184-90. Maggi's opinion is seconded by other scholars such as Lore Sander and Giuliana Martini. 4 Maggi, 'Il Grande veicolo'; idem, 'Il libro di Zambasta'; idem, 'Annotations on the Book of Zambasta. I'; idem, 'Annotations on the Book of Zambasta. II'; idem, 'Annotations on the Book of Zambasta, III'; idem, 'Annotations on the Book of Zambasta, IV'. 5 Duan, 'The Maitrī-bhāvanā-prakaraṇa', 57-107; idem, 'Zaoxiang gongde jing', 129-68. 6 Martini, 'Tracing the Sources'; idem, 'Studies on the Book of Zambasta'; idem, 'A Large Question in a Small Place'; idem, 'Bodhisattva Texts, Ideologies and Rituals in Khotan'; idem, 'The Fifth Chapter of the Book of Zambasta'; Dhammadinnā, ''Mahāratnakūṭa' Scriptures in Khotan'. 7 Chen et al., 'Mahayana Sūtras in Khotan', Hitch, 'Old Khotanese Synchronic Umlaut'; idem, 'Meter in the Old Khotanese Book of Zambasta ';idem,'tt in Old Khotanese',; idem, 'Contracted Semivowels in Old Khotanese'.
Maggi also modified parts of Emmerick's 1968 transcription and translation. 4 Apart from Maggi, Duan Qing also made corrections to Emmerick's translation; and more importantly, at Chang Lei's 常蕾 pointing, Duan discovers that Chapter 3 corresponds to the Da fangguangfo Huayan jing xiuci fen 大方廣佛花嚴經修慈分 [The Independent Chapter of the Flower Garland Sūtra on the Cultivation of Compassion] (Xiuci fen, hereafter). Furthermore, Duan re-interprets Chapter 23 based on its Chinese counterpart-the Zaoxiang gongde jing 造像功德經 [Sūtra on the Merits of Creating Statues]. 5 Giuliana Martini (a.k.a. Dhammadinnā), in her turn, traced the sources of some passages in the Book. In addition, based on Maggi's new dating, Martini uses the Book to illustrate the historical background in which Mahāyāna Buddhism emerged in Khotan, as well as to delineate the main Khotanese Buddhist beliefs and rituals. 6 Chen Ruixuan 陳瑞翾 and other scholars continue exploring Mahāyāna scriptures cited in Chapter Six. 7 Douglas A. Hitch also made a philological study of the Book. 8 Thanks to the contribution of these scholars and many others, some chapters have been studied fairly thoroughly, especially Chapters 1,2,3,5,6,13,14,17,18,and 23. The exception, however, is 9 It is traditionally believed that the Book of Zambasta originally has 25 chapters, with Chapter 25 missing, thus making Chapter 24 the last one. Maggi and Martini further suggest that the present Chapter 17 and 18 are originally one chapter, whereas the present Chapter 21 corresponds to the original Chapter 20 and 21. But this new division of chapters still does not affect Chapter 24 being the last chapter of the Book. 10 Nattier, Once Upon a Future Time, 145-227. the last chapter, Chapter 24. 9 Apart from Maggi's Italian translation, only a few passages in this chapter have been translated and properly interpreted.
The Famie Prophecy in Chapter 24 The so-called famie prophecy emerged a few hundred years after the Buddha's death. According to the prophecy, a few thousand years after the Buddha's death, the dharma will disappear due to the internal conflict of the saṃgha. This prophecy is also known as the Kauśāmbī prophecy, as it prophesizes that the dharma will end in the kingdom of Kauśāmbī. On this topic, Jan Nattier has written a monograph. In recent years, I have become interested in the evolution of the Kauśāmbī prophecy in China, Khotan and Tibet. As a story that originated in Indian Buddhism, the Kauśāmbī prophecy underwent certain transformations as it spread to Central Asia, notably Khotan, and later to China and Tibet. Chinese Buddhism, from very early on, had transformed this Indian Buddhist prophecy by adding the 10,000-year-long mofa 末法 (end of the dharma) age to the traditional 'two ages' (i.e., zhengfa 正法 (true dharma) and xiangfa 像法 (semblance dharma) ages), thereby postponing famie to an indefinitely distant future. 12 With regard to Tibetan Buddhism, it had initially accepted wholesale the Khotanese version, but eventually it also modified the story by postponing, within a certain limit, the time of famie. As for Khotanese Buddhism, even though it did not break away from the Indian tradition in terms of the storyline and finale of the prophecy, it had adapted the story in order to highlight the prominence of Khotanese Buddhism. In the Khotanese version, Khotan is portrayed as the starting place of the decline of the dharma before the dharma declines in Tibet and Gandhara and finally in Kauśāmbī. That being said, the Khotanese prophecy has remained relatively faithful to its Indian Buddhist antecedent, for in the Khotanese version, we see no trace of the mofa belief that was highly influential in East Asian Buddhism.
However, even though the present article is part of my larger interests in the origin and evolution of the famie and mofa beliefs, some findings that emerged from this investigation will be of utility also for understanding the textual history of the Candragarbha sūtra and the Book of Zambasta. And I will do so by drawing comparison between Chapter 24 and the three other versions of the prophecy of the 'Candragarbha sūtra group'; as well as between Chapter 24 and the Persian epic Shahnameh and Indian epic Mahābhārata. These comparisons will help us establish Chapter 24 as representing a new version of the Candragarbha sūtra; and contribute to determining the geographical origin and date of the Candragarbha sūtra and the Book.

Five Major Points of Discrepancies
As noted earlier, Jan Nattier has compared as many as thirteen versions of the Kausāmbī story. In this chapter, rather than making a comparison as comprehensive as Nattier, I will focus on comparing it with the texts in what I call the 'Candragarbha sūtra group' which has an either direct or indirect textual relationship with the Book. Specifically, I will focus on the Chinese Candragarbha sūtra, the Tibetan Candragarbha sūtra and the Prophecy of the Arhat Saṃghavardhana (Tib. Dgra bcom pa dge hdun hphel gyi lun bstan pa) which is the Tibetan translation of a Khotanese text influenced by the Candragarbha sūtra. After the comparison, I noticed five scenes in which the Book varies significantly from the others.

Prophecy of the Arhat Saṃghavardhana
Devas and other dharma-protectors sense that the tragic finale is approaching; the Queen Māyā summons devas and other dharma-protectors to descend to Jambudvipa.

None None None
Mara and his followers roam in the world; Jambudvipa collapses as a result.

None None None
Below I will elaborate on each point of discrepancy: 1. Regarding the scene that Duṣprasaha engages in a heated battle with the Three Evil Kings who invade Kauśāmbī and succeeds in defeating them and unifying Jambudvipa, Chapter 24 contains a detailed and vivid description of the battle, 13 whereas the other texts do not describe the battle. Some texts do not even mention the duration of the battle. Some say the battle lasts twelve years, whereas others say it is the destruction of the dharma inflicted by the Three Evil Kings that lasts twelve years, followed by a twelve-year-long period dedicated to the offerings to the saṃgha in Jambudvipa. The only other text that mentions the battle lasting three months is the Prophecy of the Arhat Saṃghavardhana which, however, does not describe the battle.
2. After the battle, Duṣprasaha invites a tripiṭaka master to expiate his sin of killing. In Chapter 24, the master reprimands Duṣprasaha for his violent deeds and preaches the precept of non-killing and non-looting, etc. 14 It is noteworthy that while all versions mention Duṣprasaha invites the tripiṭaka master, only the Book details the content of the teaching, especially the vinaya, taught by the master. 15 The Book of Zambasta ,411. 16 Ibid.,413. 17 Ibid., 419.
3. Duṣprasaha follows the master's advice and invites all saṃgha in Jambudvipa to Kauśāmbī to receive his offerings and to attend his repentance ceremonies. Due to the treacherous journey to Kauśāmbī, many die on the way, but as many as one hundred thousand saṃgha have reached Kauśāmbi. They are also the only saṃgha that remain in Jambudvipa. Here the Book makes a special note that the saṃgha who arrive in Kauśāmbi are only able to bring a small quantity of sūtras, abhidharmas and vinayas, whereas all the remaining Mahāyāna scriptures are destroyed, because all the Mahāyāna practitioners elsewhere in Jambudvipa have vanished. 15 In contrast, none of the other versions mention abhidharma, vinaya, or Mahāyāna Buddhism.
4. In Chapter 24, before saṃgha kill each other in Kauśāmbi, devas and dharma-protectors sense the imminent collapse of Buddhism; Queen Māyā, having the same pre-sentiment, summons devas and dharma-protectors to descend to Jambudvipa. 16 In other versions, devas and dharma-protectors are also featured, but it is in the context that they abandon the Three Jewels after feeling repelled by the villainous acts of the saṃgha. In contrast, only the Book describes devas and dharma-protectors as being charged by Queen Māyā to make another attempt to protect Buddhist followers. 5. After the infighting of the saṃgha and their death, the dharma disappears from Kauśāmbi. 17 As a result, Mara and his followers roam about the world and Jambudvipa collapses. This scene appears only in the Book.
Among the above five scenes, I am most interested in the description of the battle. None of the other texts describe the battle as elaborately and lively as the Book. In fact, it is quite rare to find the description of battles in Buddhist sūtras in general, for Buddhists seem hesitant to describe any warfare. For instance, the Jiading biqiu shuo danglai bian jing 迦丁比丘說當來變經 [Prophecy of Kātyāyana Bhikṣu] outright criticises the monastics who 'discuss warfare and obsess with triumphs and defeats' 或談軍馬, 鬥諍勝負. 18 It is perhaps for this reason that, except for the Book, all the other versions only provide a brief description of the battle between Kauśāmbi and the Three Evil Kings. For example, the Chinese Candragarbha sūtra writes: The three barbarian kings and their armies gradually arrive in Kauśāmbi. For twelve years, they fight. Duṣprasaha annihilates all three kings and their family, unifies Jambudvipa and becomes its supreme ruler. 彼三邊夷王, 及與諸軍眾, 漸詣拘睒彌, 十二年中鬥.
三王及眷屬, 難看王殺盡, 統領閻浮提, 而作一蓋王. 19 The Tibetan Candragarbha sūtra says the following: When twelve years have passed from the time that prince is first born, at that time the assembled armies of 300,000 of the three kings Greek, Parthian, and Saka, together with the kings [themselves], will invade the country of King Mahendrasena … then his son, the five hundred marvelous [sons] of the ministers, and so on will assemble an army of 200,000 and send it forth. And having done this, during the battle the iron mole between the eyebrows [of the prince] will grow prominent, and his entire body will become [like] iron; and filled with wrath, he will fight … then, having fought with the armies of the three kings for twelve years, [Dusprasaha] will destroy all [their] many armies. And having seized the three kings, he will put them to death. And from then on, he will become the great king of Jambudvipa. 20 21 This translation is based on Zhu's Chinese translation of the Tibetan Prophecy of the Arhat Saṃghavardhana. Zhu, 'Sengjiafatanna shouji yizhu', 476. 22 Nattier, Once Upon a Future Time, 179.
The Prophecy of the Arhat Saṃghavardhana, translated from Khotanese to Tibetan, says: The kings of three kingdoms gather and discuss their plot of invading India. At the time, in Kauśāmbi in India, there will be a king called Duṣprasaha. When he is born, the sky will rain blood; and below his elbows is red, as if smeared by blood. He will have 500 councilors with excellent physics and 200,000 valiant soldiers. Then the kings of the Stag-gig and the other two kings will enter Kauśāmbi and seek out King Duṣprasaha. After hearing this news, Duṣprasaha leads an army to wait for their arrival and subsequently fights for three months without interruption. The three pagan kings and their armies are annihilated. 21 As these passages show, all three texts contain only a limited description of the battle, but Chapter 24, in Nattier's words, shows 'good literary effects' and may have been an 'improvisation' of the compiler. 22 While it is indeed possible to attribute the battle scene to the compiler's improvisation, I wonder whether it was inspired by a source that we have not yet identified.

Possible Influences from the Shahnameh and Mahābhārata
The translation of the battle scene could be found in Chapter 24 of Emmerick's book, from the line 402 to 429, which corresponds to two folios of the original Khotanese manuscript (fol. 419 and 420). His translation is as follows: A hundred-thousand is his [Duṣprasava's] army. With his army he goes out against the kings. Their horses are groomed and harnessed with a horse's harness. They put on various armour, clothing.
Their missiles are sharp. They carry many spears. Those kings rose up hither with their army. They draw up in battle-array and begin to fight. Some bear spears against one another. They pierce very, extremely fiercely with daggers. Distraught, they attack one another with swords. Then their heads roll severally upon the ground. There is a discharge of bows, a great flashing of spears, a sound of drums as they smite with the sword. Some bear spears against one another. They pierce very, extremely fiercely with daggers. Distraught, they attack one another with swords. Then their heads roll severally upon the ground. There is a discharge of bows, a great flashing of spears, a sound of drums as they smite with the sword. Where they lie pierced, where struck, smitten, some lie lifeless. Vultures came high above them in a whirl. The earth greedily drinks their blood. The horses pierced by the vanguard rush away. They neigh loudly and run in all directions.In every place there are tumults and trouble, a loud din, cries: 'Give it to them, pierce, strike!' Flags, Makara-banners flutters about greatly. Clouds rise up. A black whirlwind comes. The fighters were struck down and many were pierced, smitten. All were smeared with blood as one sees Rākṣasas. The spears, missiles, swords in their hands gleam as lightning flashes. Very fierce is their mind toward one another. It seems very good to them when one kills another there. When the whole army from time to time sinks down, at once it rises up like the waves in the ocean. Dogs drag the dead to the bushes. Wolves, jackals howl in the distance. Some of those pierced whom they have taken have fled. Behind them are Bhūtas, strength-robbing, swift. Many fragments of arrows, spears, together with heads, lie thick on the ground in dust. With their army, the three kings retire. From them the Law disappears. The loud din dies down. Their ranks are disordered. They gallop away together in groups. They contract the far-flung regions. Scarcely is there protection for them in armour. They are afraid. Winds carry their flags backwards. No sooner does Duṣprasava come upon them with his army than these three kings are killed there. A hundred-thousand was the army of each king. They all perish. Then of the whole earth there remains at last Duṣprasava as king. 23 The Chinese name for Duṣprasaha is the 'homely king' 難看王, while the king Aśoka is also said to be ugly. One could see in Duṣprasaha other features evocative of Aśoka. This perhaps implies that the Kauśāmbi story, which has Aśoka as a protagonist, would not date earlier than the time of Aśoka. This battle scene is presented as a battle of an epic scale that determines the fate of Jambudvipa. Duṣprasava and the Three Evil Kings each lead 100,000 soldiers. In total, 400,000 soldiers engage in this bloody and dramatic battle which ends with the annihilation of the Three Evil Kings and their 300,000 soldiers, and the enthronement of Duṣprasava as the sole ruler of the world. And it is against this background that Duṣprasava invites all the saṃgha to Kauśāmbi.
Battle scenes are common in the epics around the world. In such epics, the protagonist is usually the heroic figure, around whom the battle unfolds, while the ordinary soldiers receive only sparse description and serve as no more than the foils for the protagonist. In the Book, however, Duṣprasava is not portrayed as a hero and is only described sparsely, whereas much ink is spent on depicting the dramatic battle between the two armies. Reading the text more closely, we can see that the purport of describing the battle in such dramatic details is to contrast it with a later scene when Duṣprasava repents his violence. 23 Seen from this perspective, it becomes clear that this chapter is not meant to highlight Duṣprasava's heroism, nor to describe the battle for its own sake, and much less to showcase the literary prowess of the author or the compiler. Instead, it aims to contrast Duṣprasava's violence with his subsequent repentance. And I believe the author of this scene had borrowed from the existing epics in depicting the battle.

Shahnameh
There are two epics related to Khotan that are the most likely sources for the depiction of the battle in the Book. The first is the Persian epic Shahnameh that features a battle scene involving the hero Rustam. The author of Shahnameh is Firdowsī living in the tenth and eleventh century, so the book Shahnameh itself could not have been the 24 Sims-Williams, 'Rustam and his zīn-i palang', 249-58. 25  direct source of the Book. However, the story of the battle had appeared much earlier. As far as we know, the name Rustam and similar names appeared in a third-century Greek manuscript, a fifth-century Byzantine manuscript as well as the sixth-century archaeological finds discovered in the upstream area of Indus River. We also know that in the seventh century, the legend of Rustam was already widely circulating in Sogdia. 24 Moreover, the Dunhuang Library Cave also contains Sogdian fragments that record the Rustam legend. The two fragments respectively make up the upper and lower portion of an original manuscript and are now preserved in the U.K. and France. The fragments dated to the ninth century and contain content not found in Firdowsī's Shahnameh. 25 So, it is entirely possible that the stories featured in the Shahnameh, or other heroic legends not included in the Shahnameh, had already been circulating in Persia before the Shahnameh was written. And these legends may have also spread to Khotan, considering that they are also recorded in some Dunhuang manuscripts. Besides, the earliest inhabitants in Khotan were Śaka people who spoke an Iranian language; this linguistic tie makes it more likely for Khotan to receive influences from Persian culture. 26 In addition, a wooden plank discovered in Khotan features a drawing of a male god with four arms. The god is thought to be the god of the silk embroidery and, according to Stein, shows Persian artistic traits. 27 Later scholars continue debating on this matter: some identify the god in question as a local protector god, 28 while some associate it with Jamshid, another hero in the Shahnameh. 29 Even 29 Elikhina, 'The Renovated Central Asia Exhibit in the State Hermitage Museum', 155-56. Elikhina kindly informed me that this association between the 'silk god' and Jamshid was first made by Nataliia Vasil'evna D'iakonova and was seconded by Boris Ilich Marshak. But I have not been able to identify the sources in which they made this opinion. 30 This English translation is cited from David Dick's Shahnameh.
though no consensus has been reached, there are at least speculations on a possible tie between Khotan and Persia.
The following is the scene of a battle in the Shahnameh, fought between the king Mazanderan and the king Kavus whom the hero Rustam serves: The din of drums and blare of trumpets rang out on both sides, the sky was darkened and the ground turned the color of ebony. The sparks struck by swords and maces were like lightning flashing from dark clouds, and the air was filled with the scarlet, black, and purple of banners.
The demons' cries, the darkened atmosphere, The din of drums, the horses' neighs of fear Shook the firm land-no man had seen before Such fury, or such violence, or such war; The clash of weapons filled the air, and blood Flowed from the heroes like a monstrous flood-The earth became a lake, a battleground Where waves of warriors broke and fell and drowned; Blows rained on helmets, shields, and shattered mail Like leaves whirled downward in an autumn gale. 30 There are connections to be drawn between this passage and the battle scene in Chapter 24 of the Book. All the battle elements that appear in the Shahnameh also appear in the latter, such as the general, soldier, sword, lance, arrow, banner, horse, the flood of blood, the change of color in sky. When it comes to the specific wording, however, no direct relationship could be discerned between the two 31 The passages below are cited from Kisari Mohan Ganguli's translation.
texts. In other words, they share commonalities only insofar as the elements of the battle are concerned, while these connections become less apparent when we examine the specific wording.

Mahābhārata
The second epic whose influences may have reached Khotan is the Mahābhārata. The epic unfolds around an eighteen-day battle between the Kaurava and the Pāṇḍava; and shares some common elements with the Book not found in the Shahnameh. Below I will list some of these elements. I should clarify that my intention is not to list all the shared elements, but to point out that such similarities exist between the two books. The following examples are mainly drawn from the Book of Bhishma, the Book of Drona and the Book of Karma which describe parts of the Kurukshetra War. On the left column is my summary of some passages of the Book, while on the right column are the original passages from the Mahābhārata. 31

Chapter 24 in The Book of Zambasta Mahābhārata
Soldiers on two sides wield all sorts of weapons, scream and fight to their death.
8.16 And the son of Pandu repeatedly cut off with his broad-headed shafts, steeds, riders, drivers, and flags, and bows and arrows, and arms decked with gems.
Vultures circle in the sky 7.7 And vultures and hawks and cranes and Kankas, and crows in thousands, began continually to fall upon the [Kaurava] troops. And jackals yelled aloud; and many fierce and terrible birds repeatedly wheeled to the left of thy army….
Warhorses struggle out of the rein and escaped.

Chapter 24 in The Book of Zambasta Mahābhārata
Both armies are like a succession of waves 6.119 That sea of troops, abounding in cars and elephants and steeds, and full of waves constituted by foot-soldiers, stood still as soon as it came in contact with the Satyaki continent.
7.20 And many elephants, without riders on their backs, wandered hither and thither … Wild dogs, wolfs wait their turns to feed on the corpses on the battlefield, while the frightening devils also wait for their feast.
6.50 Dogs and jackals, and crown and cranes and other carnivorous birds, and wolves and hyenas, and ravens and other blood-drinking creatures, all diverse tribes of Baksliasas, and large number of Pisachas, on the field of battle, tearing the skins of the corpses and drinking their fat, blood and marrow, began to eat their flesh. And they began to suck also the secretions of rotten corpses, while the fiakshasas laughed horribly and sang aloud, dragging dead bodies numbering thousands.
6.92 And ravens and Kankas and wolves, with great glee, drank the blood of those elephants and steeds and their Mleccha-riders overthrown on the field by Partha with his sharp shafts.
The banner on which is drawn a Capricorn.
The Mahābhārata does not feature a banner that has the symbol of the Capricorn but does mentions a banner behind Arjuna that has the symbol of a monkey-king.

Chapter 24 in The Book of Zambasta Mahābhārata
Corpses of the killed soldiers are inundated in the blood.
6.50 And that terrible river flowed fiercely through the middle of the field of battle, wafting living creatures to the regions of the dead.
6.92 Arjuna caused a fierce river to flow there whose current consisted of blood.
(Slain) foot-soldiers and steeds and cars and elephants constituted its embankments.The showers of shafts poured constituted its rafts and the hairs of the combatants formed its moss and weeds. And the fingers cut off from the arms of warriors, formed its little fishes. And that river was as awful as Death itself at the end of the Tuga. And that river of blood flowed towards the region of Yama, and the bodies of slain elephants floating on it, obstructed its current.
The killed soldiers still hold weapons tight in their hands.
7.41 The Earth has become impassable with arms still holding swords in their grasp, and with heads adorned with earrings.
Many more kindred elements could be identified between the two texts. As far as the above examples are concerned, they are all absent in the Shahnameh, but are found in both the Mahābhārata and the Book. However, in the Mahābhārata, the battle scene is merely a background, with the true focus being the heroic deeds of the protagonists. In contrast, the battle scene in the Book is elaborately portrayed with the ostensible intention of highlighting the violence and brutality of the battle. In addition, in the Mahābhārata, the passages describing the battle are scattered throughout the text, whereas in the Book, the battle scene is laid out all at once, as if the compiler had carefully handpicked sundry elements of the battle scene of the Mahābhārata in order to dramatise the battle between Duṣprasaha and the Three Evil Kings. If this is true, then the compiler must have been deeply familiar with the content of the Mahābhārata, for he had not only deliberately selected an array of literary elements from the Mahābhārata but also arranged them 32 Emmerick notices a passage at the end of Chapter Eight which is a commentary made by the compiler. Based on this passage, Emmerick believes the editor has actively edited the text. See The Book of Zambasta, 11. But in my opinion, this passage actually shows the compiler's faith in the authenticity of the sūtra, for this passage says: 'If the dharma in the Book is not accepted by the reader because it is not featured in Buddhist sūtras, then it is my fault; but if the dharma is indeed featured in the Mahāyāna sūtras, but is still not accepted for into a coherent narrative. Even certain elements seem to be deliberately selected out. For instance, the fighting of chariots and elephants which is described in great details in the Mahābhārata describes, is completely dispensed with in Chapter 24. This makes one suspect that perhaps the author is indeed a Khotanese who left out this scene in consideration of his Khotanese audience who is not familiar with the kind of battle that involves chariots and elephants.
In short, even though the compiler of the Book did not directly borrow from the Mahābhārata, he had very likely consulted the battle scene in the Mahābhārata in his effort to dramatize the battle between Duṣprasaha and the Three Evil Kings. This constitutes a unique characteristic of Chapter 24 of the Book that makes it stand out from the other versions of the Candragarbha sūtra.

Chapter 24 as Representing a New Version of the Candragarbha Sūtra
In order to establish Chapter 24 as representing a new version of the Candragarbha sūtra, we need to determine whether the battle scene was invented by the compiler based on an extant version of the Candragarbha sūtra or cited from an unknown version. I lean towards the latter possibility, because as far as we know about the Book, it was compiled by faithful Buddhist(s) who created this collection of Mahāyāna sūtras with minimum editing. Even though the compiler has appended some notes at the beginning or end of some chapters, it is unlikely that he would insert new content within the sūtra; much less likely would he have conjured up an entire battle scene. 32 its authority, then it is the fault of the reader who could not comprehend the dharma; and in this last case, the fault does not lie with me, nor with Buddha.' So, the compiler seems confident that the Book is faithful to Buddhist sūtras and doctrines. In light of this attitude, it is unlikely that the compiler would invent a scene not found in the sutras.
Not to mention that tinkering with an authoritative text is a practice frowned upon in the Indian Buddhist tradition. Therefore, it is more reasonable to suppose that the entire Chapter 24, including the battle scene, pertains to another version of the famie prophecy which the compiler included in the Book without extraneous editing, in the same way that the compiler had incorporated other texts into the Book. Admittedly, we have not identified the 'parallel texts' for all the passages of the Book, which gives room to doubt that perhaps some content was not cited but added by the compiler himself. However, more and more 'parallel texts' have been identified, which shows the Book as a collection of texts with identifiable sources. As for the minor discrepancies between the Book and the sources, they could be considered the common corollaries of any translation process. In short, given the overall style of compilation of the Book, it is highly probable that Chapter 24 is also borrowed entirely from a non-extant version of the Candragarbha sūtra.
Besides, earlier I have established five major points of differences between Chapter 24 and the other three versions (i.e., Chinese and Tibetan Candragarbha sūtra, and the Prophecy of the Arhat Saṃghavardhana). In particular, the last point of difference, regarding the elaborate battle scene in Chapter 24, is a significant structural difference rather than minor differences in wording. sūtra have been discovered. 33 Generally speaking, even though the earliest Sanskrit manuscript emerged as early as the third century, the majority of Sanskrit manuscripts do not predate the fifth or sixth century. 34 Furthermore, by observing the script of the Sanskrit fragments which resembles that of the Gilgit manuscripts, scholars have suggested that these fragments could not date earlier than the sixth century. 35 This means that even though these Sanskrit fragments are the earliest extant version of the Candragarbha sūtra, this version does not necessarily reflect the original composition of the Candragarbha sūtra.

Geographical Origin and Dating of the Candragarbha Sūtra
What makes the Candragarbha sūtra less likely an Indian work is the fact that the 'Xingsu sheshou pin' 星宿攝受品 [Chapter on Constellations] of the Candragarbha sūtra mentions twenty-eight constellations, which clearly denotes an influence of Chinese astrology. 36 This is the result of either Narendrayaśas who, in translating the Chinese Candragarbha sūtra, modified it based on his knowledge of Chinese astrology; or the result of the influence of Chinese astrology in Khotan. If the latter were the case, the Candragarbha sūtra could be determined to be a Khotanese work. Certainly, these questions 37 Narendrayaśas actually translated the Candragarbha sūtra before the Sūryagarbha sūtra, even though the Sūryagarbha sūtra is the earlier work. 38 Lévi, 'Notes chinoises sur l'Inde'. 39 Yu, 'Songyun, Huisheng xishi de ruogan wenti', 63-100.
could not be answered with certainty unless we find Sanskrit or Khotanese counterparts to this chapter, but in any case, it is clear that such content could not have appeared in an Indian Buddhist text. Sylvain Levy also determined the Candragarbha sūtra to be a Khotanese Buddhist work. He made this conjecture on the ground that the Sūryagarbha sūtra, a text closely connected with the Candragarbha sūtra, is a Khotanese Buddhist text. The Candragarbha sūtra starts with 'Having taught the Sūryagarbha sūtra' (說 《日藏 經》 已), suggesting that the Candragarbha sūtra is sequential to the Sūryagarbha sūtra. 37 Moreover, Lévy points out that the Candragarbha sūtra, like the Sūryagarbha sūtra, also shows numerous connections with Khotanese Buddhism. All these observations led Lévy to conclude that the Candragarbha sūtra originated in Khotan.
The Khotanese origin of the Candragarbha sūtra is plausible also because the sūtra likely originated to the east-and not to the west-of the Pamir Mountains. Because if it had originated to the west, then Narendrayaśas, as someone from Oddiyana, should have been familiar with the text. However, Narendrayaśas's translations of both the Candragarbha sūtra and Sūryagarbha sūtra are often ambiguous, suggesting he is unfamiliar with the sūtra and Khotanese Buddhism in general. 38 This is understandable, as Narendrayaśas had never been to Khotan and came to China through northern prairies and not by the Northern or Southern Silk Road. Besides, Narendrayaśas apparently did not carry scriptures with him during this long journey to China. 39 So, Narendrayaśas must have translated the manuscript of the Candragarbha sūtra that was obtained by the Northern Qi regime from the Western Regions, and perhaps directly from Khotan.
Regarding the date of the Candragarbha sūtra, Lévy observes that the kingdoms of the Western Regions mentioned in the 'Fenbu Yanfuti pin' 分佈閻浮提品 [Chapter on the Spread (of the Dharma) 40 Thomas,Tibetan Literary Texts,45. in Jambudvīpa] of the Chinese Candragarbha sūtra reflect the geo-political situation of the fourth century. Based on Lévy's opinion, F. W. Thomas simply considers the Candragarbha sūtra as a fourth-century Khotanese work. 40 But the fourth century clearly can only be the terminus ante quem of the sūtra. In addition, considering that Narendrayaśas (517-589) translated the Candragarbha sūtra into Chinese in Yecheng in 566, the Candragarbha sūtra must have been circulating in Khotan between the fourth and mid-sixth century.
Another clue for dating is found in a statement in Chapter 24 that says the dharma will last 1,500 years after the Buddha's death. In my upcoming book, I argue that before the late fifth century, it was not the '1,500 years', but the '1,000 years' that was the common length of time assigned to the duration of the zhengfa and xiangfa ages. This fact also makes it unlikely that the Candragarbha sūtra was composed before the late fifth century. By the way, the Chinese Candragarbha sūtra also adopts the '1,500 years' time frame, whereas the Tibetan Candragarbha sūtra says 2,000 years. This means the Chinese Candragarbha sūtra and the Book can be dated closer in time, whereas the Tibetan Candragarbha sūtra should be dated later. It is even possible that the Chinese Candragarbha sūtra and the Book share the same prototypical text, namely, a Khotanese Candragarbha sūtra.
Finally, if the connection between the Mahābhārata and Chapter 24 could be established, then the Mahābhārata can also help us indirectly infer the date of the Candragarbha sūtra. As we know, the Mahābhārata was composed and compiled through a long period of time from the fourth century BCE to the fourth century CE. This gives two possibilities for the Book to be influenced by the Mahābhārata: either in the unfixed form of the Mahābhārata that spread to Khotan before the fourth century or in its fixed form after the fourth century. Judging by the similarities of the battle scene, I think it is more likely that the Khotanese Buddhists had received the Mahābhārata in its fixed form and thereafter incorporated its literary elements into the Khotanese literary repertoire, before using them to 41 The Book of Zambasta,363;Nattier,'Church Language and Vernacular Language',210. create the Candragarbha sūtra. And in my opinion, it requires of a culture a relatively significant amount of time to 'absorb' a foreign text before it could use these foreign elements creatively for the sake of its own literary production. If there is any validity in this speculation, then the fifth century becomes an unlikely date, for in this case, Khotanese would only have a century or less between their reception of the Mahābhārata-which was compiled in the fourth centuryand the writing of the Candragarbha sūtra. In short, I believe the early sixth century is the most likely time range for the composition of the Candragarbha sūtra.
In summary, both the Candragarbha sūtra and Sūryagarbha sūtra show Khotanese indigenous features incongruous with the Indian tradition. However, these evidence alone do not suffice to support the conclusion that the two sūtras were written by Khotanese in the Khotanese language, because it is also possible that the indigenous features were added by the Khotanese compilers. However, its description of the famie story, especially the battle scene, is more than a simple 'tinkering'. Besides, the Book, as far as we know, is compiled with minimum changes made to the original texts; and this applies to Chapter 24. All these evidences considered, and despite the fact that only the Sanskrit, not Khotanese manuscripts of the Candragarbha sūtra have been found, we should not haste to rule out the possibility that it was originally written in Khotanese. This hypothesis needs further corroboration and depends especially on the future discovery of relevant manuscripts. But it is certain that Chapter 24 of the Book is highly significant for our study of the Candragarbha sūtra.

Dating the Book of Zambasta
So far, several dates have been suggested for the Book. The most common opinion holds that the Book was composed no earlier than the seventh century. 41 For instance, Skjӕrvø suggests the Book exem- 42 Skjӕrvø suggests that Khotnaese underwent three stages of development: the Early Khotanese (fifth-sixth c.), the Middle Khotanese (seventh-eighth c.) and the Late Khotanese (ninth-tenth c.). See Skjӕrvø, 'Khotan, An Early Center of Buddhism in Chinese Turkestan'. As for whether this classification should be maintained, it is debated by the Khotanese specialists. 43 Maggi,'Khotanese Literature',. This opinion, however, is somewhat at odds with the current scholarly consensus that the kingdoms around the Tarim Basin did not start using their languages to translate Buddhist texts until the very end of the fifth century or the sixth century. For this reason, we should examine Maggi's conclusion with a more critical pair of eyes. plifies the 'Middle Khotanese' from the seventh to eighth century. 42 Maggi, on the other hand, suggests the Book of Zambasta dated to the late fifth century. Based on his dating of the Book and other paleographical evidence, Maggi draws a broader conclusion that Khotanese Buddhist texts had appeared as early as the second half of the fifth century. 43 Personally, I think Maggi's date is too early. Since I am not an expert on paleography or history of languages, I could not challenge him in these aspects, but I do want to turn to the content of the text for some clues.
My first counterargument lies with the Candragarbha sūtra which, as I argued earlier, is likely an early sixth-century work; and since the Book has cited the Candragarbha sūtra, it must have dated later than the early sixth century.
Chapter 23  Khotanese. In this case, the 'I' in the epigraph must refer to the translator who translated the Zaoxiang gongde jing into Khotanese, whom the compiler quotes directly. Other scholars believe the 'I' is the compiler himself who not only compiled the Book, but also translated this chapter. Also see Duan, 'Zaoxiang gongde jing ', 129-68. it is very agreeable, but they so learn it in Kashmirian that they also understand the meaning of it. To the Khotanese that seems to be the Law whose meaning they do not understand at all.... In words the essential thing is the meaning.... The meaning being unperceived, no one would escape from the woes in saṃsāra. 44 This epigraph above tells us that Khotanese Buddhists at the time valued Indian sūtras more than Khotanese translations despite their limited knowledge of Indian languages; and for this reason, the epigraph cites Chinese and Jibin Buddhists as examples to convince Khotanese Buddhists that the Khotanese translation of Indian Buddhist sūtras is more conducive to learning and understanding. An important implication from this epigraph is that some Indian sūtras had already been translated to Khotanese and were being circulated to the extent that Khotanese texts were competing with Indian originals. Such situation is unlikely to have happened as early as the late fifth century. In fact, many Mahāyāna texts included in the Book seem to have already been translated to Khotanese by the time the Book was compiled, suggesting that the Book had emerged during a mature stage of Khotanese Buddhism.
Third, if the Book had indeed been translated to Khotanese in the second half of the fifth century, then we should have seen a translation of lower calibre. After all, the earliest translations in a new language tend to be poor in quality, which, however, is not what we observe with the Book, suggesting that the Book does not represent the earliest Khotanese translation.
Fourth, the Book draws from sources written in Khotanese, Sanskrit, Central Asian languages, and even Chinese. 45 This reflects that by the time when the Book was compiled, many Indian and Central Asian Buddhist texts, and a sizeable number of Khotanese texts, had already been in circulation. 46 Such level of Buddhist textual exchanges is unlikely to have occured as early as the late fifth century, but much more likely if we date the Book later.
Lastly, the sources of the Book also help us determine the terminus post quem. For instance, some passages of the Pusa dichi jing 菩薩地持經 [Skt. *Bodhisattva-bhūmi; Holding of the Bodhisattva Grounds], a late fourth-century text, appear in the Book. However, while it is tempting to use this fact to set the fifth century as the terminus post quem, these passages are also found in the Da baoji jing 大寶積經 [Skt. *Mahāratnakūṭa sūtra; Sūtra of the Heap of Jewels] which is a much earlier text. This makes the Pusa dichi jing no longer a particularly useful source. 47 In comparison, the Xiuci fen, which Duan Qing identifies as the Chinese parallel of Chapter 3, is more helpful in this regard. This discovery puts into question the 'fifth century' dating, because if the Xiuci fen had indeed been translated to Khotanese in the fifth century, then it is difficult to explain why it was not translated to Chinese until the late seventh century, especially given the intense cultural exchanges between the two countries.

Conclusion
By way of a conclusion, I want to say a few more words on the connection between the Book and the Mahābhārata. This connection is made on the basis of the battle scene found in Chapter 24 of the Book. As for whether such connection is meaningful, or whether it 48 It is worth mentioning that in the Dunhuang Library Cave were discovered the Khotanese manuscripts of another Indian epic, the Rāmāyaṇa (P.2801, P.2781, P.2783). The three manuscripts are supposed to make up one single juan, but were broken into three. The content diverges significantly from the Sanskrit version, but is closer to Tibetan translation. See Rong et al., Yutian yu Dunhuang, elucidates any issue, I am receptive to any criticism.
Personally, I believe this connection could be explored further. First, the similarities between the Mahābhārata and the famie prophecy are not limited to the battle scene. For instance, several versions of the famie prophecy describe the Duṣprasaha as wearing an armor at the time of his birth, while in the Mahābhārata, the hero Karna is also born with an armor (1.57.82). Similarly, the famie prophecy in Chapter 24 ends with a quarrel within the saṃgha on the night of the precept meeting (Skt. poṣadha) which leads to the bloodshed and death of the entire saṃgha; similarly, in the Mahābhārata, after the victory of the Kurukshetra War, as Krishna and his clan Yadava celebrate the victory, an internal strife leads to the extinction of the entire clan. Such similarities do not necessarily mean the famie prophecy has directly borrowed from the Mahābhārata, but the similarities are so intriguing that one suspect whether there are indeed connections to be made between the non-Buddhist traditions in India and the Indian Buddhist famie prophecy. 48 Secondly, it is perhaps possible to refine Emmerick's English translation of the Book by consulting the Mahābhārata. For instance, in Emmerick's translation, the bodies of the dead soldiers 'were smeared with blood as one sees Rākṣasas'. This translation feels somewhat forced. However, by consulting the Mahābhārata, we could render the translation as 'Rākṣasas are waiting to feed on the bodies of the dead soldiers.' Lastly, as far as the battle scene is concerned, Chapter 24 shows more influences from the Indian Mahābhārata than from the Persian Shahnameh, suggesting a stronger influence from India than from Persia. However, this does not mean the Indian culture had exclusive influences in Khotan. In fact, the artefacts discovered in Khotan show non-Indian features. After all, Khotan, as an important station on the Southern Silk Road, has absorbed influences from a gamut of cultures, notably the Persian and Chinese culture. More research could be done to study the relative degree of influences that Khotan had received from the Persian and Indian culture, as well as the intense exchanges between China and Khotan.